An intelligent lung nodule segmentation framework for early detection of lung cancer using an optimized deep neural system

Lung cancer is the most dangerous disease in the world, leading to high mortality in daily life. So, the early detection of this disease is essential to enhance the survival rate of patients worldwide. However, the small extent of lung nodules is not easily found by eye vision, so this leads to the inaccurate diagnosis of lung cancer. Nowadays, deep neural systems technology-based machine learning has been widely used in healthcare units to diagnose various diseases. It is one of the best ways to classify subtle parts of a CT scan exactly. Therefore, proposed a Sailfish-based Yolo Segmentation Framework (SbYSF) that aimed to segment the lung nodule part accurately. Initially, the CT image datasets were collected preprocessed, and required features were extracted using the sailfish function. The nodule region is traced through the extracted features and segmented. Further, the severity is attained using the SbYSF framework. The planned model accurately detected the nodule region for the early stage of a lung cancer diagnosis. The proposed model is checked in the Python simulation environment. They reached the highest accuracy, precision, and recall of 99.75% and F-measure of 99%.